中文摘要：

流域水文模型由于受到众多因素的影响导致结构复杂、参数众多，大多数参数都有明确的物理意义，理论上可根据其物理意义直接定量，但实际上由于缺乏观测数据支持，往往需要通过系统识别的方法推求。因此，积极开展模型参数识别扳术的研究和应用从而得到合适的参数对模型推广应用及提高应用精度都有很重要的意义。采用由粒子群算法和混沌优化方法结合的混沌粒子群算法对水文模型参数进行优选，该算法对局部最优解在搜索空间上进行混沌迭代优化，改善和提高了基本粒子群算法的全局寻优性能和收敛速度。在新安江水文模型参数优选的应用结果表明，该参数优选方法比传统参数优选方法更易快速的收敛于全局最优解。

英文摘要：

The hydrological models have a complicated structure and lots of parameters because of many factors. Most of the parameters have explicit physical significance, which can be measured directly according to their physical significance. But we always obtain them by system recognition because we have not observed data. So we should get the reasonable parameters by initiatively carrying out the research and application of the technology in parameter recognition, which is significant to the promotion and application of the models. To optirlize the parameters of the hydrological model, this paper has taken chaotic particle swarm optimization which combines partical swarm optimization with the chaotic optimization. This optimization takes the chaotic optimization on the local extremum in the space, has enhanced the capability of precise search and improved the speed of the global search. The application of chaotic particle swarm optimization to the parameter optimization of Xinanjiang model indicates that to achieve the globe extremum, this parameter optim：zation is fatter than traditional parameter optimization.